Project Summary Low weight eating disorders are severe psychiatric disorders that most often develop in adolescence, have a chronic course, and evidence poor response to treatment. A core feature of these disorders is food avoidance; which can be conceptualized in the research domain criteria (RDoC) framework as dysregulation in the dimensional expression of acute threat (i.e., food aversion) and reward learning (i.e., pleasurable response to food absence). Emerging neuroimaging evidence suggests that deficits in insula-amygdala-ventral striatum (IAVS) neurocircuitry contribute to individual variability in aversive and reward learning, and that these brain regions demonstrate abnormal responses to food/eating stimuli. Our pilot data suggest that patients with low weight eating disorder (LW-ED) experience difficulty extinguishing food-cue associations in a reversal learning paradigm compared to healthy controls, a difficulty that is related to psychophysiological measures of aversive- disgust (not fear). Additionally, our data indicate the presence of a pleasure response to food avoidance or ?cued safety? that may further impair reversal of food-cue associations because it prevents reappraisal of the threatening stimuli. We have also successfully piloted an interoceptive exposure intervention for this population that targets visceral sensitivity and seeks to increase ?top-down? regulation of the IAVS neurocircuit. This intervention is distinguishable from existing contingency management based behavioral interventions that are standard for this population and target the reward value of food and eating. Further, simultaneously acquired fMRI-electromyography(EMG) pilot data from our team suggests that an elevated aversive-disgust response is associated with increased connectivity between insula and amygdala, and the cued absence of the food reward is associated with elevated connectivity between amygdala and ventral striatum that is associated with elevated pleasure response. The proposed project will (a) use novel fMRI-EMG to test the relationship between effective connectivity within amygdala-insula-ventral striatum network and its relationship to psychophysiological and behavioral measures of acute threat and reward learning in 60 adolescents with LW- EDs and 30 healthy controls, (b) test the sensitivity of this network to an experimental interoceptive exposure paradigm relative to patients receiving family based therapy for weight restoration using dynamic causal modeling of fMRI-EMG data pre-post experimental conditions, (c) validate this model against objective measures of laboratory and real world eating behavior. The results of this study will help us better understand the core neurocircuitry that underlies both threat processing and reward/aversive learning and how this circuit relates to objective behavior. Further, we will determine the modifiability of this neurocircuitry via two distinct behavioral interventions chosen to target different aspects of affective processing and reward learning. These data will be used to inform future clinical interventions targeting aversive/reward learning within this population and dysregulation in insula-amygdala-ventral striatum subcircuits.